Numbers, Facts and Trends Shaping Your World

Third-party and independent candidates for president often fall short of early polling numbers

Independent presidential candidate Robert F. Kennedy Jr. speaks during a voter rally in Aurora, Colorado, on May 19, 2024. (Helen H. Richardson/MediaNews Group/The Denver Post via Getty Images)
Independent presidential candidate Robert F. Kennedy Jr. speaks during a voter rally in Aurora, Colorado, on May 19, 2024. (Helen H. Richardson/MediaNews Group/The Denver Post via Getty Images)

The 2024 presidential campaign stands out as the first presumptive rematch between major-party candidates since 1956. It’s also the first time an ex-president has run to reclaim the White House in more than a century.

Another uncommon feature is the presence of several high-profile alternative candidates, including Democratic-scion-turned-independent Robert F. Kennedy Jr., independent Cornel West and three-time Green Party nominee Jill Stein.

Kennedy, an environmental lawyer and anti-vaccine activist, is currently polling in the mid-single digits nationally. He appears to draw support both from people who might otherwise back President Joe Biden and former President Donald Trump, complicating both men’s campaign calculations. (Bear in mind that accurately gauging support for third-party candidates can be tricky.)

But U.S. political history tells us that third-party and independent candidates usually finish a lot lower than where they start.

We examined preelection polls in six presidential contests that featured significant third-party or independent candidates, then reviewed those candidates’ actual shares of the popular vote in the general election.

Not only did support for third-party and independent candidates tend to decline over the course of their campaigns, but their vote shares often came in lower than polls suggested they might.

Here’s an election-by-election look at underperformance by third-party and independent candidates.

How we did this

Given the unusual dynamics of the 2024 presidential election – including the presence of several potentially significant third-party and independent candidates – Pew Research Center examined how such candidates fared in past elections.

We focused on the six elections over the past 60 years in which the major-party share of the nationwide popular vote was less than 98%. In each of those elections, an independent or third-party candidate won at least 2% of the vote.

For each of those candidates, we obtained support-level data via iPoll, an online archive of historical survey data maintained by Cornell University’s Roper Center for Public Opinion Research. For 1980 and subsequent elections, we limited our analysis to surveys of registered voters. No such surveys were available for the 1968 election, so in that case we used surveys of the national adult population.

Over the decades, survey modes shifted from predominantly face-to-face interviews to landline telephone interviews, and then to landline-plus-cellphone interviews. By 2016, online surveys were making their first appearances, but most polls were still conducted via phone. To avoid any distortions caused by such different survey modes, we used only surveys conducted by the same mode within a given year. This meant that we only used face-to-face surveys in 1968, and only phone surveys in all other years we analyzed.

We also looked at the wording of each individual question to make sure each survey was asking essentially the same thing in similar ways. In particular, we wanted to ensure that candidates were referred to by name and identified by party (or as “independent” when appropriate).

Once we had assembled a list of comparable questions, we plotted support for third-party and independent candidates on a timeline. The final point on each chart represents the candidate’s share of the total nationwide popular vote. For 1968 through 2000, we used figures from America Votes, a long-running compilation of election data. For the 2016 election, we compiled official returns from all 50 states and the District of Columbia.

With two exceptions, all support figures in this analysis include those who said they would vote for or leaned toward the candidate in question. The exceptions are John Anderson in 1980 (because no surveys with “leaner” questions met our inclusion criteria) and Ross Perot in 1992, during the interim period in which he wasn’t actively campaigning (because surveys did not typically ask “leaner” questions about him during this period).

1968: George Wallace

A scatter plot showing support for George Wallace in 1968.

Fresh off his first term as Alabama’s segregationist governor, George Wallace – running a “law and order”-themed campaign under the American Independent Party banner – saw his support rise in polls over the spring and summer leading up to the 1968 election. In April, around 10% of adults nationally said they supported or leaned toward Wallace. By September, that had doubled to 20%. Wallace appeared within reach of his goal: dividing the field enough to throw the election to the House of Representatives, where he could try to bargain his electoral votes for “concessions” on desegregation, voting rights and other issues.

That fall, Republican Richard Nixon’s campaign began warning conservatives that voting for Wallace would only help Democrat Hubert Humphrey. Meanwhile, Democratic-aligned unions worked to pull their members – whom Wallace had targeted – back into Humphrey’s fold. Wallace’s running mate, retired Air Force Gen. Curtis LeMay, also made headlines at his introductory press conference after saying he’d consider using nuclear weapons in Vietnam.

Wallace’s support in the polls began to slide, reaching the mid-teens in the weeks before Election Day. He ended up with 13.5% of the popular vote and 46 electoral votes – not enough to keep Nixon from winning the White House.

1980: John Anderson

A scatter plot showing support for John Anderson in 1980.

Rep. John Anderson of Illinois was trailing badly in the Republican presidential primaries when, in April 1980, he dropped out and said he would run as an independent instead. Anderson’s candidacy generated considerable public interest: Around 20% of registered voters said they would support him, and he continued to poll around that level throughout the spring.

But Anderson’s nascent campaign had to spend much time and energy that spring and summer simply getting his name on state ballots. Anderson faded from view during that summer’s Democratic and Republican conventions. Incumbent President Jimmy Carter, the Democrat, refused to share a debate stage with him in the fall – though Republican nominee Ronald Reagan did debate Anderson one-on-one.

By October, Anderson’s support in polls had dwindled to the 9%-10% range. In the end, he won 6.6% of the national popular vote.

1992: Ross Perot

A scatter plot showing support for Ross Perot in 1992.

Money and visibility weren’t issues for Ross Perot, the billionaire businessman from Texas who mounted a stop-and-go independent campaign against Republican President George H.W. Bush and his Democratic challenger, Arkansas Gov. Bill Clinton.

Perot’s effort, driven initially by volunteers and appearances on Larry King Live, quickly gained momentum. In March, as Perot’s backers began gathering the hundreds of thousands of petition signatures he would need to get on state ballots, Perot was regularly receiving support from 20% or more of registered voters in polls. By May, about a third of registered voters were telling pollsters they’d vote for or were leaning toward Perot. In a few surveys, he led both Bush and Clinton.

Amid sharpening attacks from Republicans and Democrats, though, Perot’s numbers began falling. In mid-July, when his support was below 20% in most polls, Perot abruptly quit the race.

Although Perot was no longer actively campaigning, his name remained on two dozen state ballots, and some never-say-die supporters continued working to gain him ballot access in additional states. Pollsters continued to ask voters about Perot throughout the summer and fall – especially as speculation grew that he might jump back into the race. While Perot’s support declined steadily during this interim period, in late September around 10% of voters still said they preferred him to Bush or Clinton.

Perot reentered the campaign in early October, and within a few weeks his support had climbed back up to around 20%, including leaners. It began to slip again as Election Day neared, falling to around 15%. In the end, Perot won 18.9% of the popular vote – the best showing by a non-major-party candidate since Theodore Roosevelt 80 years earlier.

1996: Ross Perot

A scatter plot showing support for Ross Perot in 1996.

Perot wouldn’t come close to that in his second campaign. At the start of the year, when it was still unclear whether he would seek the nomination of the Reform Party (which he had founded the year before), his support among registered voters typically was in the mid-teens.

But Perot’s support declined during the campaign, eventually settling at around 5%-7%, including leaners. His poll numbers did pick up a bit in the run-up to Election Day, when he received 8.4% of the popular vote. Among the minor candidates Perot beat out for third place: consumer advocate Ralph Nader, who took 0.7% representing the Green Party.

2000: Ralph Nader and Pat Buchanan

Nader had a considerably higher profile four years later, when he was again the Green Party’s nominee. Polls taken during that close, contentious campaign regularly found that around 5% of registered voters said they supported or leaned toward Nader.

A scatter plot showing support for Ralph Nader in 2000.

That was enough to concern Democrats that Nader threatened Vice President Al Gore’s chances of defeating Republican Texas Gov. George W. Bush. (Whether he in fact did so is still hotly debated among political scientists, journalists and other observers.)

In the end, Nader won only 2.7% of the national popular vote. But in several closely divided states – including Florida and New Hampshire, both of which Bush carried – Nader’s share was enough to potentially swing the outcome.

Another third-party candidate in 2000 received a fair amount of public and media attention: Pat Buchanan, the conservative commentator who had captured the nomination of Perot’s Reform Party. Buchanan polled as high as 4% in the spring, but by fall was mostly in the 1%-2% range. He ended up with less than 0.5% of the popular vote, but did well enough in five states to theoretically (or perhaps not so theoretically) affect the outcome.

2016: Gary Johnson and Jill Stein

A scatter plot showing support for Gary Johnson in 2016.

Widespread dissatisfaction with Republican Trump and his Democratic opponent, Hillary Clinton, may have caused more voters than usual to look beyond the major parties. Two candidates in particular received considerable attention: former New Mexico Gov. Gary Johnson – the Libertarian Party nominee – and physician and activist Jill Stein of the Green Party. (Both Johnson and Stein had also run in 2012, though with less impact.)

Johnson polled fairly strongly into the fall, with 8%-12% of registered voters routinely saying that they would vote for him or were leaning toward him. But Johnson’s poll numbers began trending downward, and by Election Day his support level was hovering around 5%-6%. Johnson ended up receiving 3.3% of the vote – the 52-year-old Libertarian Party’s best showing in a presidential election to date.

For her part, Stein often received support from 5%-7% of registered voters in polls taken during the spring and summer of 2016. But her support also eroded as the campaign went on, and she eventually received just over 1% of the popular vote – still the party’s best result since Nader in 2000.

  翻译: